Intelligent Things – From Prompts to Ghosts

Hello everyone,

Welcome to the “Intelligent Things” webinar hosted by ALBA.

I would like to take this opportunity to acknowledge the ongoing work of Mr. Samer Eid in introducing and developing alternative ways of thinking through the use of new tools for design and expression.

I would also like to thank Dr. Ricardo Mbarkho, Director of ALBA’s Center for Research and Creation, who is hosting this initiative, encouraging it, and integrating it into the framework of academic inquiry and rigor.

Thank you to our six international speakers who, from the Far East to the Far West, will each share their interests, valuable knowledge, and ambitions, and to Gilles Chahwan for moderating these interventions.

Finally, thank you to Dr. Gaby Issa El Khoury, Director of the ALBA School of Architecture, for his encouragement and unwavering support for the education and personal growth of our students.

Artificial intelligence (AI) is part of a long history of celebrating the human mind. Long before circuits and neural networks, the Persian poet Ferdowsi celebrated reason in the 10th century as humanity’s treasure and compass, reminding us that wisdom illuminates action and elevates peoples. This poetic insight foreshadows the idea that intelligence – whether natural or artificial – is not merely an accumulation of facts, but the art of organizing, discerning, and anticipating.

A century later, another Persian, the mathematician al-Khwarizmi, laid the technical foundation for this endeavor. By systematizing algebra and computational methods, he bequeathed “algorithms” – a term derived from his name – that is, finite procedures for solving problems. Without these operational frameworks, there would be no programming, no machine learning, and no engines capable of extracting patterns from data. AI thus continues an age-old endeavor: transforming mathematical abstraction into operational power.

On a philosophical level, Descartes’s “I think, therefore I am” reminds us that thought is experienced through its own activity. AI, for its part, does not “think” like a conscious subject; it calculates, models, and anticipates. But the cogito invites us to question what we call thinking, and to distinguish between lived experience and functional competence.

Spinoza, finally, sees at the heart of life a dynamic of growth and perseverance in being. This idea inspires an ethics of AI: designing systems that augment our capabilities, promote the common good, and integrate harmoniously into our environments. At the intersection of poetry, mathematics, and philosophy, AI thus becomes a project of lucidity, method, and responsibility. It contributes to the expansion of our understanding of reality, from science to art and emotions.

Architecture and Artificial Intelligence: The True Potential of Artificial Intelligence

Ladies and Gentlemen,
Speakers and panelists,
Dear colleagues and fellow educators,
Dear students,

Today, we find ourselves at a defining turning point in higher education – and even more so in architectural education. Artificial Intelligence has already disrupted the academic landscape. We can no longer ignore its presence, nor its use – at times excessive – among students. The question, therefore, is no longer whether AI should be part of our teaching, but how we choose to engage with it.

Higher education has already weathered technological “revolutions” that challenged our pedagogical methods. Allow me to briefly mention two examples: the calculator, which replaced the slide rule and allowed us to go further in the study of sciences; and Computer Aided Design, CAD, which transformed – for students and professionals alike – the way we work and share design drawings.

You might argue that these two revolutions pale in comparison to the one brought about by AI – and you would be right. Yet in their own time, they were experienced as real shocks within the world of education. What feels unprecedented today often echoes transformations we have already learned to absorb.

The greatest fear stems from one central question: can AI replace the architect? This concern is understandable, because unlike previous revolutions, AI possesses a creative potential that those earlier disruptions simply did not.

Through Large Language Models, LLM, students can now watch their ideas transform into objects – what we call AI-Generated Content, AIGC. Hence the impression that the architect might become redundant. And yet, this impression deserves to be questioned. It is precisely the architect who feeds data into these models, and it is still the architect who critiques the outcomes of that intelligence. The student, like the professional, remains the final decision-maker – accepting or rejecting what the machine has created.

And this is where it gets fascinating. Who actually created that form? The human, through their analysis and the prompts they entered? Or the machine, which interpreted that input in its own way? Without claiming to settle the debate, I would say spontaneously: a little of both.

Educators tend to criticize their students for their overreliance on AI, while students have not yet mastered the relationship between their inputs and the outcomes they receive. What AI saves in time should allow students and professionals to reinvest that time – in refining their ideas, personalizing them, and truly making them their own. Used well, AI does not diminish authorship; it challenges us to redefine it. What we perceive as the exclusive work of AI should instead be understood and experienced as a close collaboration between AI and the user – whether student or professional – while acknowledging the control that humans must maintain over the machine.

Acknowledging the benefits of AI, universities in general, and ALBA in particular, can no longer afford to ignore this technology: they must integrate it into their curricula as swiftly as possible. This is not only a matter of keeping pace with technological change; it is a condition for remaining intellectually relevant. Not only is it necessary to step fully into the 21st century and its new technologies, but we can no longer hope to attract the next generation of students without them.

I now invite you to enjoy this webinar: Intelligent Things. The title is revealing in itself – things, by nature inert and devoid of any emotion, have today become intelligent.

Six speakers will guide you through this journey:

• Dr. Lev Manovich, Professor at the City University of New York, CUNY, will show that AI tools can “reason” and are capable of understanding and interpreting images.
• Dr. Cesare Battelli, Professor at Shaoxing Vocational and Technical College, offers a different perspective on AI, conceiving it as a contemporary heterotopia – real, yet existing in “another” space.
• Dr. Matias del Campo, Associate Professor at the Technische Universität Wien, TUW, will argue that AI-assisted design is a precise theory, not a matter of chance.
• Karel Klein, Senior Lecturer at the Southern California Institute of Architecture, SCI-Arc, reassures us that AI does not annihilate our ability to perceive and reproduce objects – on the contrary, it strengthens it.
• Yara Feghali, Lecturer at the University of California at Los Angeles, UCLA, and an ALBA alumna – which I could not fail to mention – offers what amounts to a practical catalogue of ways to work with these new tools.
• Ebrahim Poustinchi, Associate Professor at Kent State University, KSU, award-winning architect, and founder of the Robotically Augmented Design Lab, RAD, will close this webinar by integrating robotics and Augmented and Virtual Reality, AR/VR, into everything that will have come before.

I leave you now to enjoy, this afternoon, these six talks – which I am convinced will open new horizons – and new ways of working and thinking about space, in architecture and beyond.

Good evening, everyone,

Since its establishment in 2023, the Centre for Research and Creation (The CRC of ALBA) has been working to support research and creation through a strong framework, critical reflection, and interdisciplinary exchange. The Centre was created as a place where artistic disciplines can meet one another, but also enter into dialogue with other fields of knowledge.

This is important because the arts do not stand outside society. They are part of it. They reveal its tensions, its transformations, its contradictions, and sometimes even its blind spots. Across history, artistic research and artistic practice have often questioned what a society takes for granted. They have challenged norms, opened new ways of seeing, and invited people to think differently.

Our role is to create conditions for questioning, experimentation, and critical engagement. At the CRC, uncertainty, doubt, and reflection are part of the process. They help us resist passive reception and encourage more active forms of thought and practice.

In this sense, the artist-researcher is not just someone who produces work. He or she also develops a position toward the world. The social and cultural environment becomes a field of inquiry, a source of hypotheses, and a site of intervention.

Because the CRC is based in Beirut, it gives special attention to the evolution of the creative sectors in Lebanon and to their social, cultural, and political implications. At the same time, it remains open to exchanges with international partners, research centres, and labs. This dialogue across contexts is part of our vision.

Tonight’s webinar fully belongs to this approach.

INTELLIGENT THINGS invites us to reflect on artificial intelligence as a transformation that is already affecting how we write, imagine, design, represent, and think. This concerns architecture and many other fields. It also concerns education, research, and the future of creative work itself.

The question, then, is not whether artificial intelligence is good or bad in itself. The real question is: what does it change in our relation to images, language, authorship, and thought? What happens to the act of creation when generative systems begin to participate in the production of forms? What happens to the role of the author, to the status of the project, and to the place of critical judgment?

These are not small questions. They ask us to think carefully, and also collectively.

This is why such a webinar matters. It creates a shared space for reflection. It brings together voices, approaches, and disciplines around an issue that is already reshaping many parts of our intellectual and creative lives. For instance, in the Lebanese context, it also raises questions related to security, surveillance, and the broader conditions within which technology is deployed.

I would like to warmly thank all those who helped make this event possible, as well as our speakers for joining us and sharing their work and ideas. I also thank the audience.

With our guests, let us expand the space for thought and discuss new paths for research and creation.

Thank you, and welcome.

« Prompts ignite in the midnight grid. Hosts decode what the black glass hid. Ghosts survive in the spectral stream. And the circuit folds into a dream. Intelligent Things. Shadows that think. In the walls. In the wires. Built from ghosts and desires – We live among Intelligent Things. » [ I*T © 2026 ]

Good afternoon and welcome to all.

“INTELLIGENT THINGS | from prompts to ghosts” – is not just one catchy theme; it is a genuine posture. The latter finds solid anchor within a simple postulation; a subtle displacement in which ‘intelligence’ is no longer where we thought it was. It no longer resides in the exclusive author, nor in the tool, but circulates and fluctuates – fragmented and redistributed – across systems that think with us, beside us, and sometimes ahead of us. In this sense, the current initiative does not aim to instruct in usage, but to situate creative practices amongst broader grounds, where the boundaries between human and machine, the conceived and the computed, grow increasingly porous. A trilogy scheme feeds upon such infrastructure: three intertwined domains, statuses or key-figures; sequentially labeled – ‘Prompts’. ‘Hosts’. ‘Ghosts’. This dynamic triptych inflects synchronously through linguistic, operative, and spectral entanglements.

Prompts serve as point of entry. Accurate commands or triggering vectors, they establish directions, gradually enabling the models to interpolate, extrapolate, and generate throughout drifting workflows. To prompt is to engage with the very intricate grammar of the text, corresponding syntax and context. Cognitive constructions emerge from explicit instructions, performative gestures refract into embodiment, therefore mediating between language connotation and algorithmic depiction, in quest of nascent meaning. The prompt engineer becomes a designer of inputs, a strategist of iteration; a curator of trajectories.

At the core of this evolving paradigm – Hosts surface as active vessels. If prompts initiate processes, hosts sustain and materialize them. They are the simulated environments within which agency is enacted, channeling abstract computation into perceptible artefacts, intelligently mapping things and territories. The augmented user, aided by means of highly dexterous engines and resourceful datasets, glide and slide across diffusion signals and scenes, strata of co-production, bridging thereby the relational gap between human and nonhuman protagonists. Spatial protocols deploy, crystallizing towards speculative ecologies.

Ghosts are the ambient forces and hidden traces, underneath latent worlds. Echoes of expired typologies, layers of mnemonic imagery; residual patterns and segmented agents, recursive biases and memorial cues; they condense the intangible pedigrees of forms, with contorted genealogies. Not glitches or anomalies. Uncanny outputs thus manifest in atmospheric modulations, opaque tendencies, distortions and intensities; implicitly encrypted, via subconscious renditions. Sedimented altogether in a metadata-driven repository of accumulated histories, mute diagrammatic simulacra haunt cartographies of post-digital archaeologies.

That was for the narrative, in brief.

Before we proceed, I would like to extend few words of gratitude – to the ALBA Dean, Prof. Fadlo Dagher; to Dr. Gaby Issa-El-Khoury, Director of the School of Architecture; to Dr. Ricardo Mbarkho, Director of the Centre for Research and Creation, and co-organizer of this event; one final mention goes to Gilles.

Dear ALBA students and colleagues, this webinar offers us a theoretical premise, a curatorial stance, and most importantly – the lucid realization that those “INTELLIGENT THINGS” already affect our futures. Thank you all for being here, despite the challenging circumstances – Now, I hand over to Dr. Mbarkho.

Good afternoon,

Introducing Lev Manovich means situating a voice that has impacted how we think about media for more than three decades.

He is Presidential Professor at the Graduate Center of the City University of New York. His background moves across painting, architecture, filmmaking, and computation. He began working with digital media in the early 1980s, long before it became a dominant field, and has been teaching it since the early 1990s.

Across this trajectory, he has contributed to the formation of several key areas: new media studies, software studies, cultural analytics, and more recently, the study of AI aesthetics. His work is widely read and translated, with hundreds of publications and a series of influential books, including The Language of New Media, Software Takes Command, and more recently Artificial Aesthetics.

What is important here is not only the quantity of work, but its position. His writing does not simply describe technological change. It proposes frameworks to understand how culture itself is reorganized through media. His artistic work follows the same direction. His projects have been exhibited in major institutions such as the Centre Pompidou and the Shanghai Biennale. This double position, as both theorist and practitioner, gives his work a particular clarity.

In his lecture, A Medium That Thinks, Manovich proposes a shift that concerns all of us.

He suggests that generative AI is not just another tool. It behaves as a medium that can process, interpret, and generate content with a form of internal logic. It draws from vast histories of images and styles, and produces outputs as continuous variations rather than identical copies. This has direct consequences.

Authorship becomes less stable. The role of the creator changes. The work is no longer fully controlled from the beginning to the end. It emerges through an interaction.

Within this context, the prompt takes on a new role. It opens a process in motion without fully defining what will emerge from it.

What is at stake is a new relationship between thinking and making. Media ceases to be a mere channel. It becomes an active part of thought itself.

Lev Manovich, welcome.

We now turn to Cesare Battelli.

Cesare Battelli is an Italian architect and researcher working at the intersection of architecture and artificial intelligence. His work develops through projects, teaching, and a long-term research platform called Visionary Architecture, which he founded in 2003.

He studied architecture at IUAV in Venice, developed work in Moscow, and later completed a Master’s at the Städelschule in Frankfurt under Enric Miralles and Peter Cook. He then obtained a PhD in architecture in Spain. His career includes teaching, workshops, and juries at institutions such as Politecnico di Milano, the Architectural Association, and other international schools. His projects and writings have been published in journals like El Croquis, Architectural Design, and Frame, as well as widely shared on digital platforms. He is currently active as a researcher and professor, continuing to develop work that moves between theory and speculative design. What defines his position is a clear shift in how architecture is understood.

In his lecture, Heterotopias, Battelli proposes to read AI less as a tool for efficiency than as a space. A space where different times, forms, and possibilities coexist. This idea comes from the notion of heterotopia. A space that contains multiple realities at once, without reducing them to a single order. Within this framework, AI becomes a place where architecture moves beyond conventional design. It appears, transforms, and sometimes resists definition.

The role of the prompt changes here as well. It is no longer a command producing a fixed result. It acts as a threshold, a point of passage between language and image.

The outputs that emerge from AI carry a specific quality. They are not fully stable objects. They appear as traces, as forms that suggest other possible worlds without fully resolving them. This leads to another shift.

Architecture is no longer only about building. It becomes a practice of emergence. The architect composes and navigates within a field of possibilities.

Placed next to Manovich, this position extends the discussion. From a medium that thinks, we move toward a space where worlds take shape.

Cesare Battelli, welcome.

If architecture appears mainly as image, what does it become?

Can it still be understood as architecture when it circulates more through renders, screens, and visual speculation than through construction? Where do we draw the line between architecture and visual fiction?

What does the algorithm change in the way we think and imagine?

Does it reveal new possibilities, or does it impose its own logic on what can be produced? Is AI opening imagination, or narrowing it through repetition and prediction?

We entered a site – inferential or interferential?

A comprehensive field where images, forms and spatial structures seamlessly merge with statistical memory. Therein, from tool to milieu, Artificial Intelligence infiltrates the architectural discipline, re-informing its epistemological scope. Dr. Matias del Campo thoroughly cultivates within this threshold.

Trained under the eminent Hans Hollein at the University of Applied Arts Vienna, and later completing his doctorate at RMIT Melbourne, Dr. del Campo is both practitioner and theorist. Co-founder of SPAN Architecture, his pioneering office spans across Vienna and Shanghai, combining advanced technology and philosophical inquiry, with a focus on “voluptuous data”, machine learning and AI-driven design methods. Professor at TU Wien’s Institute for Digital Architecture and Planning, and CEO of Angewandte Künstliche Intelligenz, he consistently maneuvers at the confluence of speculative research and operative intelligence systems.

Matias del Campo resists consuming AI as mere automation, reframing it instead as a cultural condition, in which intelligence is embedded, distributed, and always in formation. Precedent is no longer referenced in the linear historical sense, but compressed and encoded into high-dimensional spaces, reanimated as latent possibility. Within this framework, models are not basic means of output, but epistemic devices that reshape how architectonics are rationalized, assembled and remembered. Hence, the morphological act develops less as a routine of calculation, more as a process of solicitation revealing summoning momenta; relentless negotiations with vast, invisible archives of built and unbuilt worlds, where every product carries traces of innumerable lineages, transmuting toward a relational lattice of destabilized artefacts beyond recognition. What emerges is neither solely authored, nor entirely autonomous, but disperses in probabilistic synthesis across human and machine, inside a complex continuum of computational becoming.

Recipient of the Accelerate@CERN Award [2015] and the ACADIA Innovative Research Award of Excellence [2024], del Campo’s work is held in institutions like the MAK Vienna, the Pinakothek München and the MAXXI Rome. Resolutely constructed over series of projects along with pedagogical experiments, Dr Matias del Campo’s compendium includes robust publications such as: “Sublime Bodies” [2017], “Machine Hallucinations” [2022], “Neural Architecture” [2023], and “Diffusions in Architecture” [2024], with “The Architects Manual to Artificial Intelligence” forthcoming. An ontological apparatus for understanding generative protocols, not only as language patterns, but as active intellectual forces that constantly rescript the multilayered affiliation between instrument and environment via prior knowledge. Computation converted to conjuration.

A stitched passage from Jean Baudrillard’s “Simulacra and Simulation” [1981] enhances the previous equation:

« Simulation is no longer that of a territory, a referential being, or a substance, but the generation by models of a real without origin or reality: a hyperreal. The simulacrum is never what hides the truth; the simulacrum is true. It is no longer a question of imitation, duplication, nor even parody. There is no longer any referential: the real has been abolished, leaving only its operational double. The territory no longer precedes the map, nor survives it – henceforth, it is the map that precedes the territory. »

Through “The Séance of Inference” – Dr. del Campo traces a dynamic route, from evocation to invocation. Here, the prompt becomes an incantation; the medium, a host for odd elucubrations; extensive ecologies of ghostly manifestation. We entered this site – at once interferential and inferential – flowing with substantial apparitions, endowed with multiple registers. We are no longer alone in the deeds of making. Something else insists, suggests, deviates. Something, altered. Something, displaced. Something, other. An irreversible entanglement, where the human is no longer distinct from the posthuman it has beckoned. Now, we all gather – to listen to what is inferred.

We lost something – not intelligence, but attunement.

The ability to sense emotional resonance before reason, to read the world as signal rather than statement. Modernity traded that faculty for clarity. Artificial Intelligence, strangely, tends to give it back: misaligned, unstable, unreadable, yet, propitious. Karel Klein furtively operates inside such a rupture.

Academically trained on engineering precision and architectural speculation – beginning at the University of Illinois and sharpened at Columbia’s GSAPP during its early 90s’ “paperless studio”, Klein’s trajectory has continuously moved along a fault line tangled in between structure and imagination, system and drift. Today, as co-director of Ruy Klein in Los Angeles and a faculty member of SCI-Arc (Southern California Institute of Architecture) and WashU (Washington University in St. Louis), her practice unfolds across design, pedagogy and research, as a constant investigation of how avant-garde architecture might think.

Karel Klein does not use AI to optimize form or accelerate production. She misuses it, deliberately. Since 2016, her alchemical experiments have relentlessly trained models on ‘things’ architecture is not: myriads of items, trivial instruments and anatomies – funneled into re-enchanted architectural bodies. What emerges is a field of estranged image-objects in limbo; all too familiar in trace, alien in effect. There, displaced memories flicker via metaphysical transfers, not as references, but as reincarnations.

Formerly hosted within (mis)trained styleGANs (Generative Adversarial Networks) then more recently based on LoRA-driven datasets (Low-Rank Adaptation), morphological assets acquire material textures, behaviors and atmospheres of novel worldbuilding. A digital cabinet of curiosities, deliriously reconfigured with a menagerie of mutant taxonomies, ornamental artefacts, exquisite crossbreeds, and spectral chimeras, narrating new myths-in-becoming. Shapeshifting encounters of a ritualistic kind, “they are the frozen moments of deconsecration, the surrealist games of automatic (automated) writing; spontaneous and meaningless, disturbing and enchanting”. Quoting Steve Baker’s excerpt from “The Postmodern Animal” [2000]:

« The object – I cannot give it a better name – was discovered in a dimmed corner of a room filled with other objects one might simply call ‘curiosities’. This object had many visible qualities of an animal but was assuredly not one. Eyes without a face. Musculature beneath tufted fur, yet wooden, intricately carved limbs. Furniture-like arms and legs. The purpose of the object was unclear. It appeared useless. I wondered about its meaning. Was it also meaningless? I did not know this object’s identity nor its provenance. I decided to take this object with me to learn its story and failing that, to invent one. »

Chronologically exhibited at SCI-Arc Gallery [2017], FRAC Centre d’Orléans [2019] and the Venice Biennale [2021], Klein’s works are progressively supported by a series of theoretical essays including “Verto Pellis” [2020], “Machines are Braver than Art” [2022], “Machines à Rechercher” [2022], “To Think a New Thing: AI, Metaphor and the Fantasies of Knowing” [2023], and a current book in the making, titled “Reading What Was Never Written: Nonsensuous Similarities in Artificial Intelligence and Divinatory Practice”. Her AI-corpus articulates an adamant aesthetic – as a cognitive engine for immersive scenographies, rendering unmapped realities.

Klein’s lecture “Lost Faculties and Estranged Matter” – to be performed shortly within our ALBA webinar “Intelligent Things | from prompts to ghosts” – navigates and probes this ambiguous terrain: intelligence, not as property but as migration; not authorship, but distributed agency; not vision, but a recalibration of perception itself. We lost something – not intelligence, but attunement. What returns is not what was lost. It is stranger than that. A faculty that does not explain the world but makes it shimmer again. A prolific machination built on error and excess, where architecture is no longer composed, but divined.

Now, we are all eyes and ears for the goddess to sing.

Opening frame

I would like to begin by expressing my deep admiration for both of you – not only in terms of intellectual acknowledgment, but for the impact across academic contexts and within the international avant-garde. What is especially compelling is how each of your practices engages AI not just as a toolset, but as a way of rethinking what architecture can be. We have heard two distinct positions: one framing AI as a synthesis of latent systems; the other evolving as a site of misuse, distortion and perceptual attunement. Both of you seem to suggest that something elemental has shifted: not just how we design, but what it means to discern, to enquire, and to conceive. I would like to articulate our next conversation around three key-notions: intelligence, pedagogy and collateral ‘things’ – in order to reflect on the points where your trajectories converge and where they quietly diverge – with one central question in mind: what should we be teaching the next generation of architects in this condition? Ready for the discussion?

Core question #1 [ intelligence ]

• A common kick-start question: How do you each define intelligence in your respective practices, especially when distributed between human cognition and computational systems? Consequently, how could we understand intelligence today when learning – or teaching within arts and architecture?
• A direct follow-up: Where do you locate the momentum when intelligence comes to be architectural?

Core question #2 [ pedagogical speculation ]

• What could architecture education become in tandem with AI – not just as content, but as structure, ethics, and purpose? What should education focus on the most: outcomes, processes, or systems literacy? What are the essential skills, the new fundamentals – students need to acquire nowadays?
• What role should experimentation, ambiguity, or intuition play in education – should these be protected or persecuted? How can academic institutions boost this devoid of sabotage or limitation?
• How do you teach students to remain critical, while working with those inherently opaque systems?
• Do your approaches suggest shared pedagogical strategies, do they reinforce each other, or are they categorically pointing toward drastically opposed design cultures, or architectural futures?
• What should we actually stop teaching, or – put inversely – what maybe should students unlearn?

Core question #3 [ collateral things ]

• Here comes the unavoidable cliche: is authorship still a meaningful concept, or should it be promptly replaced with distributed agency? Is mastery about maintaining coherence or instability? Is the human figure still central to architecture or just one agent among many? Pushing further: if something problematic emerges from our system, who is accountable? The machine, no one or us?
• Are we still designing objects, or have we toppled from now on, into petty manipulators of systems that breed variations of objects? Are we still accessing knowledge or producing it retroactively? Rephrasing: Does our role suddenly swing from author to curator of system behavior? How do we ensure that students do not fall into passive users of engines, but survive as prolific designers?
• A technical question I lately ask myself: is AI eventually bringing the death of former software?

Closing frame

What emanates from this conversation is ultimately not a fixed model nor a single position, but a set of tensions and contingencies, trapped between synthesis and misalignment, inference and attunement, system and deviation, control and emergence. Perhaps the most significant task for future architects is not how to resolve these tensions – but to learn how to juggle with them. Karel, Matias, thank you both.